Abstract Tuberculosis (TB) is a leading cause of morbidity and mortality in Zambia, especially among HIV-infected persons. The lack of adequate diagnostic tools is a key challenge to TB control in the public health system. The Xpert MTB/RIF platform has the potential to transform TB diagnostics by providing accurate same-day results, but studies are required to evaluate the feasibility and cost-effectiveness of this technology before it can be scaled-up. We propose an evaluation of Xpert MTB/RIF using a quasi-experimental before-after study design in pediatric and adult HIV-infected populations in two similar peri-urban district hospitals in Zambia. We will compare the standard algorithmic approach to TB diagnosis endorsed by the Zambian Ministry of Health to an enhanced algorithm that incorporates Xpert MTB/RIF. The primary outcomes will be the proportions of adult and pediatric patients receiving appropriate TB treatment in each study phase, using mycobacterial culture as the reference standard, and the feasibility and cost effectiveness of Xpert MTB/RIF in this setting. Secondary outcomes include a comparison of the time to diagnosis and treatment initiation in TB-infected patients, a comparison of clinical outcomes at three and six months post TB screening between the two study phases, and an evaluation of the diagnostic performance of Xpert MTB/RIF using sputum and gastric aspirate specimens in the pediatric population. This study will enable the Ministry of Health to make informed decisions regarding future implementation of Xpert MTB/RIF and intensified case finding in Zambia. 2